Since you can buy modern servers with multiple CPUs, large amount of memory, lots of storage, and high-bandwidth network connections, each OS can have as much computing power at need as an individual server might provide. The operating systems and applications seldom use all available resources, especially at the same times–so more operating systems and applications can coexist on a single piece of hardware, resulting in better utilization of hardware.
If you’re not using virtualization now, you should be thinking about it. Here’s why.
1. It enables you to get the most out of your server hardware.
Many servers typically run for most of the day at very low levels of utilization. Adding multiple OSs and applications helps you get the most out of those expensive server resources. Since many virtual machines can run on one server, you substantially reduce the total expenditure for hardware.
2. It will allow you to set up redundant servers for better fault tolerance.
3. Management is simplified.
Since all the operating systems on a hypervisor can be administered through a single interface, and systems can be given more computing resources as needed, it can be much easier to manage a dozen virtual servers than a dozen separate physical systems. Each operating system must still be managed separately, but add-on tools are available to simplify that process as well.
4. You can partition applications on separate OSs for greater reliability.
Normally, a physical server will run one operating system, and that OS will have multiple server applications installed on it–a Web server, email server, database server, and so forth. However, running multiple server applications on a single OS increases the possibility that one application may interfere with another, causing bottlenecks or even crashes. By partitioning applications in different virtual machines, each application has its own OS and resources, and is less likely to interfere with other applications.
5. Provisioning new servers for prototyping, testing and migration is simplified.
With virtualization, creating an additional server for testing takes a few minutes, and doesn’t require any additional hardware. In contrast, buying a new physical server (or keeping extras on hand for testing) is expensive, and installing the operating system and applications can be quite time-consuming. Since existing VMs can be cloned with all operating system configurations and applications, it’s a very simple process duplicate a production system to test new patches or a new version of an application.
6. You can save energy.
7. Virtual Desktop Infrastructure can save you a fortune on PCs.
In addition to virtualizing server OSs and applications, you may want to consider virtualizing desktops. VDI allows you to run a thin client on existing desktop hardware or new thin client hardware, and access a virtual Windows 7 or other desktop operating system. This means that you can provide Windows 7 desktops to users whose existing systems will not support Windows 7 running locally. Management can also be simplified, since patches to Windows and applications can be applied once and take effect for every system on the VDI server.
There’s a reason virtualization is being used so extensively through most corporate data centers. It offers compelling savings in hardware, energy usage, management costs, and it supports great fault tolerance.